# Session Init

> Lightweight session initialization with context recovery

- **Type:** Skill
- **Install:** `agentstack add skill-mshadmanrahman-pm-pilot-session-init`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [mshadmanrahman](https://agentstack.voostack.com/s/mshadmanrahman)
- **Installs:** 0
- **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [mshadmanrahman](https://github.com/mshadmanrahman)
- **Source:** https://github.com/mshadmanrahman/pm-pilot/tree/main/skills/productivity/session-init
- **Website:** https://github.com/mshadmanrahman/pm-pilot#readme

## Install

```sh
agentstack add skill-mshadmanrahman-pm-pilot-session-init
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Session Init

Quick-start a session by loading prior context and surfacing open work.

## When to Use

- Starting a new session
- Returning after a break
- Picking up someone else's work

## Procedure

1. **Load Memory**
   - Read `MEMORY.md` (auto-loaded, but scan for recent entries)
   - Note any `currentDate` or session-specific flags

2. **Check Handoffs**
   - Read latest file in `.claude/handoffs/` (sort by date desc)
   - Extract: last task, decisions, blockers, next steps

3. **Check Open Tasks**
   - Read `tasks/todo.md`
   - Count open vs completed items
   - Identify highest-priority incomplete item

4. **Check Lessons**
   - Scan `tasks/lessons.md` for entries from last 3 sessions
   - Note any recurring patterns or warnings

5. **Output Summary**
   Produce exactly 5 lines:
   ```
   Session: {date}
   Last: {what was last worked on}
   Open: {count} tasks ({top priority item})
   Blockers: {any blockers or "none"}
   Next: {recommended first action}
   ```

## Output Format

```
Session: 2026-03-14
Last: Lead scoring simulator deploy
Open: 3 tasks (fix CPL calculation)
Blockers: none
Next: Run verification loop on simulator changes
```

## Rules

- Never spend more than 30 seconds on init
- If no handoff exists, say so and ask user for context
- Do not auto-start work; present summary and wait for direction

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [mshadmanrahman](https://github.com/mshadmanrahman)
- **Source:** [mshadmanrahman/pm-pilot](https://github.com/mshadmanrahman/pm-pilot)
- **License:** MIT
- **Homepage:** https://github.com/mshadmanrahman/pm-pilot#readme

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-mshadmanrahman-pm-pilot-session-init
- Seller: https://agentstack.voostack.com/s/mshadmanrahman
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.